MegEngine

What it does

MegEngine is an open-source deep-learning framework that combines training and inference. It supports model quantization, dynamic shapes and preprocessing, with memory-management and execution features intended to reduce the resources required to run models.

Developers use Python for training and can deploy inference across supported CPU, GPU and mobile environments. Platform support differs between Python installation and inference deployment. The repository documents older Python and operating-system requirements, so check the chosen release and build instructions before adopting it for a new environment.

What it can do

• Model training
• Inference deployment
• Quantization
• Dynamic shapes
• Memory optimization
• CPU and GPU execution

Before you choose

• Training and inference have different platform support.
• The documented Python installation range is 3.6–3.9; verify current release compatibility.
• Compute hardware and hosting are separate costs.

Listed under

• Coding
• Sandboxes and infrastructure

Get started

1. Check supported Python, operating-system and accelerator versions.
2. Install a compatible wheel or build from source.
3. Train or import a model and validate the target inference deployment.

MegEngine is available under Apache License 2.0. Compute, hardware and operational costs remain separate.

We couldn’t find published prices for MegEngine. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 30 Oct 2026. Each source supports part of the overview; availability and prices can change.

• GitHub - MegEngine/MegEngine: MegEngine 是一个快速、可拓展、易于使用且支持自动求导的深度学习框架 · GitHub ↗github.com

Company website

github.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Human review

Checked by a person

knowledge_base

manual_review

Last checked 30 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.